Trusted application of optee make boot problem as systemd service

Robotics & Edge Computing Jetson & Embedded Systems Jetson Xavier NX
,
1

hi i made a trusted application for jetson xavier nx that does very simple task to sending a parameter to trusted world and get another parameter from it in rich world and show it, and its work fine when i run it in simple way in terminal with command
after that i want to this application run as a service after every boot, so i made this service
[Unit]
Description=test_application
After=nv-tee-supplicant.service
Requires=nv-tee-supplicant.service

[Service]
ExecStart=/usr/sbin/optee_custom_example
Type=oneshot
User=root

[Install]
WantedBy=multi-user.target
but there is a problem after all of things that need to done to the service be enable when i restart the jetson the jetson dosent boot completely it will restart it rapidly some times it will show me some message that show the kernel panic happened
[end kernel panic - not syncing: oops: fatal exception]
what is the problem ?? is it because of configs of my service or optee make systemd lock somehow to if any changing happend in systemd system sont boot up ??
i will be so thankful if you help me

3

hello rabanimisagh,

may I confirm which Jetpack release version you’re now using.
since you’re develop your own TA, did you also rebuild ATF and OP-TEE from sources, and updating binary accordingly?
besides, you may also gather the complete error logs for reference.

4

im using kernel version is 5.10.104-tegra
jetpack 5.1 and l4t 35.2.1
and yes i rebuild atf and all binaries of optee sources need to be rebuild to new application work
and application works normally when i run it in terminal after complete boot
and about the error logs
as system dosnt boot to i be able copy the error logs i take a picture from monitor i hope be helpful

IMG_20240808_122013
IMG_20240808_1220131920×1582 462 KB

5

There is no update from you for a period, assuming this is not an issue any more.
Hence we are closing this topic. If need further support, please open a new one.
Thanks

hello rabanimisagh,

please have issue narrow down,
did you confirm this is reproduced with your own TA? since your failure looks like NULL pointer dereference from GPU side.
or… may I confirm your steps in details to reproduce this failure.